Hardware Stop Errors?

[Auric]

Problem: stop errors, primarily 7f, after running trouble-free for around 18 months with mostly the same hardware (problem remains with new HW removed).


Type: Error
Source: System Error
Event ID: 1003

Error code 1000007f, parameter1 00000008, parameter2 80042000, parameter3 00000000, parameter4 00000000

-
Type: Information
Source: Save Dump
Event ID: 1000

The computer has rebooted from a bugcheck. The bugcheck was: 0x0000007f (0x00000008, 0x80042000, 0x00000000, 0x00000000)

-
Type: Information
Source: Save Dump
Event ID: 1001

The computer has rebooted from a bugcheck. The bugcheck was: 0x0000008e (0xc0000005, 0x00000000, 0xf9ef8368, 0x00000000)



Investigation:
-analyzed dump files (dumpchk/pstat/windbg) inconclusive, generally pointing to ntkrnl.exe (or ntkrnlmp.exe)
-temperature is not a problem
-tested RAM and in diff'rent slots
-tried Safe Mode, old OS images and new OS setups (both XP and 2000) and flashed with various BIOS versions and generally disabled all CMOS options thus isolating to hardware
-removed or swapped all hardware except SMPS, CPU, mainboard; power is never lost (always dumps memory to HDD when so configured) so I reckon it's down to CPU or mainboard

Peculiarities:
-failure never occurs in DOS (memtest, ghost, etc), only Windows (or Windows part of OS setup)
-failure can otherwise happen at anytime but can occur several times in a row during boot attempts, each time during Windows loading, then once Windows is loaded, may not occur for some time
-savedump system event is always logged even when boot appears to complete without interuption

I do not have an alternate CPU or mainboard to test. What are the chances that either went bad in such a way that DOS seems unaffected? Is there a DOS utility that could thoroughly test the CPU and report errors? If not, I will have to choose one to swap... CPU is easiest but I wouldn't mind ending up with the latest version of the same mainboard and all the better if that was the fix, while I prefer to stick with the same model CPU.

Any suggestions would be appreciated.

 

[mechBgon]

Can you add the full system specs to that info What is an SMPS, by the way?

 

[ScrapSilicon]

Originally posted by: mechBgon
Can you add the full system specs to that info What is an SMPS, by the way?

switching mode power suppy perhaps.. smps
Anyway this is interesting and M$ just put it up last week.
There are other hits on google about your apparent " double panic " ( not my way of putting it..just how its called in the google results) error. This is all speculation on my part but I'll say test your proc in another mobo if possible. Goodluck.

 

[Auric]

Specs are in sig. I did see that MS article. Microcode updates are ruled out 'cause OS and BIOS are the latest and greatest and in any case the CPU did work fine with them previously. Environment is out too... as I type CPU is running at defaults and temp is 28C.

Unfortunately I just don't have 'nutha CPU or system to cross-test and will prolly have to buy a CPU. But I don't think the same model (which I prefer for higher FSB) is available locally anymore and I think neither the CPU nor the mainboard are under warranty any longer and it seems the mainboard could be replaced anyway because the later BIOS' (required for micro code updates and to run later CPU's) hang when integrated USB is enabled. I knew about this hang thing ages ago but not the workaround until now and had resigned meself to an earlier BIOS.

So it looks like there is no quick 'n' easy fix for the whole mess. Not that I want to just throw money at it, but even that won't do as there is the time factor.

 

[Auric]

Hmm, I was thinking about throwing a cheap Celeron in there to test... see Celeron D 2.4 fits the bill cheapwise and recalled seeing D support in Gigabyte's BIOS list... then I see that the D is actually Prescott core based which I already communicated with Gigabyte aboot as they inititally claimed and then withdrew compatiblity claims for the my Rev. 1.0 board. ... double check BIOS list ... yeah, but it says check CPU compatibility list ... okay... so that list confirms not only Celeron D but also P4 Prescott? Yet not my Northwood 2.4C (HT / 800 FSB) that it has always ran??? Gigabyte es muy loco!

I wonder if a Celeron D will really work....

 

[mechBgon]

So what voltage and timings is your RAM set up for? Also, heretic that I am, I am going to suggest that you re-couple your hard drive as a fact-finding step, if you mean that it's hanging on rubber bands or something.

 

[Auric]

I appreciate your help. Voltage and timings were defaults for testing... also tried upping the voltage. Tried diff'rent HDD's "uncoupled".

Slapped a C2.4-D in there and no stop errors so far and I have inititated oodles of manual reboots and cold boots. However, a savedump system event persists, despite a lack of symptoms...

Type: Information
Source: Save Dump
Event ID: 1000

The computer has rebooted from a bugcheck. The bugcheck was: 0x0000007f (0x00000008, 0x80042000, 0x00000000, 0x00000000). A full dump was not saved.



...but during my extensive troubleshooting travails I did notice a lack of same savedump event on an earlier image (whilst different stop errors continued) and at one point it seemed related to a patch KB885884 because they could be eliminated by removing it. On the currently running image, said patch refuses to uninstall so I still have to at least try my "basic XP image" and/or do a new OS setup to confirm.


Good news, kinda, is that I discovered that the integrated USB is not really fooked with the later Celeron-D/Prescott supporting BIOS' because it does work with Gigabyte's supplied port bracket with a mouse attached... the boot only hangs when my USB card reader is attached (it still works as ever once in Windows). ah heck.. but card readers are cheap to replace... but I have to wonder if the same problem could occur with another.


As far as Gigabyte's support list, I guess it is just not comprehensive or is it wrong showing a Northwood 2.4 800 FSB sans HT?


Intel P4-Northwood 3.4G (HT) 800 OK OK
Intel P4-Northwood 3.2G (HT) 800 OK OK
Intel P4-Northwood 3.0G (HT) 800 OK OK
Intel P4-Northwood 2.8G (HT) 800 OK OK
Intel P4-Northwood 2.6G (HT) 800 OK OK
Intel P4-Northwood 2.6G 800 OK OK
Intel P4-Northwood 2.4G 800 OK OK


Anyhoo, the way its looking now I prolly will not have to replace the mainboard and will have to see whether the 2.4C can be replaced under warranty or not and if not will order another 'cause even though I apparently have the Prescott option now I suppose the 2.4C is still the best bang for the rupee given this system.

 

[Auric]

The savedump upon every boot was bogus. I'll put it down to an OS defect since it could be "unstuck" by changing the 'Startup and Recovery' - 'Write debugging information' option. I happened to change if from Kernel to Small memory dump. Changing it back to Kernel did not cause the savedump to reoccur.

-
Type: Information
Source: Save Dump
Event ID: 1000

The computer has rebooted from a bugcheck. The bugcheck was: 0x0000007f (0x00000008, 0x80042000, 0x00000000, 0x00000000). A full dump was not saved.
-

As for the USB hang with later BIOS... its the fault of my card reader, not the integrated USB or BIOS in general -although since it didn't happen with early BIOS' I guess partial blame could be there.

So, problem solved. Bad CPU. Just wish I had had more parts ('nutha CPU) laying about to resolve this in minutes rather than days.

Thanks for listening.